Author Topic: Jim's HLV Spacecraft Design Thread  (Read 88171 times)

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 28616
  • Minnesota
  • Liked: 8585
  • Likes Given: 5600
Re: Jim's HLV Spacecraft Design Thread
« Reply #220 on: 02/15/2012 05:39 PM »
There are also manufacturing reasons to use a non-monolithic design for the primary.
Chris  Whoever loves correction loves knowledge, but he who hates reproof is stupid.

To the maximum extent practicable, the Federal Government shall plan missions to accommodate the space transportation services capabilities of United States commercial providers. US law http://goo.gl/YZYNt0

Offline go4mars

  • Senior Member
  • *****
  • Posts: 3752
  • Earth
  • Liked: 152
  • Likes Given: 3156
Re: Jim's HLV Spacecraft Design Thread
« Reply #221 on: 02/15/2012 05:56 PM »
As I stated in other thread, a bigger fairing could have been done for 500M or less (probably much less).

If greater payload volume is so beneficial why has a 6 or 7m payload fairing not been developed for the EELV’s?
They tried that.  Look up Hammerhead fairings.  Caused severe buffeting during flight.
Elasmotherium; hurlyburly Doggerlandic Jentilak steeds insouciantly gallop in viridescent taiga, eluding deluginal Burckle's abyssal excavation.

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 28616
  • Minnesota
  • Liked: 8585
  • Likes Given: 5600
Re: Jim's HLV Spacecraft Design Thread
« Reply #222 on: 02/15/2012 06:03 PM »
As I stated in other thread, a bigger fairing could have been done for 500M or less (probably much less).

If greater payload volume is so beneficial why has a 6 or 7m payload fairing not been developed for the EELV’s?
They tried that.  Look up Hammerhead fairings.  Caused severe buffeting during flight.

Respectfully, Downix can't be used as a source. ULA still offers the large fairings in their user manuals for EELVs, so that beats speculation by amateurs.
Chris  Whoever loves correction loves knowledge, but he who hates reproof is stupid.

To the maximum extent practicable, the Federal Government shall plan missions to accommodate the space transportation services capabilities of United States commercial providers. US law http://goo.gl/YZYNt0

Offline go4mars

  • Senior Member
  • *****
  • Posts: 3752
  • Earth
  • Liked: 152
  • Likes Given: 3156
Re: Jim's HLV Spacecraft Design Thread
« Reply #223 on: 02/15/2012 06:22 PM »
ULA still offers the large fairings in their user manuals for EELVs, so that beats speculation by amateurs.
I'll accept that.  So you and baldusi are implying that if NASA thinks Ares V would have saved 30% on JWST development costs (2.55 billion+) then presumably a mega-hammerhead would have saved that too.  So why wouldn't they just do a hammerhead and EELV with a monolithic scope?  Maybe:

1)  Untested hammerhead fairing experiment on multi-billion dollar payload = undesireable.

2)  A 7.2 meter fairing might still be too small to justify telescope design changes (end up with a slightly bigger complicated umbrella instead of a simpler bigger cheaper one). 

3)  They would have to pay for the upgrade out of this project's budget (which might not be possible given money management rules, i.e. only the fairing upgrade justification department can pay for it for them, but that budget is already spoken for or too small)  I don't know if this is the case or not, but I've worked in big enough companies to see how silly things like that can get in the way (drilling departments with way to much power).

4)  The monolith in a hammerhead on EELV option might have weighed too much.   

Please feel free to validate or destroy these 4 guesses with data.  If they are valid, then we come back to the premise of the thread. 
« Last Edit: 02/15/2012 06:29 PM by go4mars »
Elasmotherium; hurlyburly Doggerlandic Jentilak steeds insouciantly gallop in viridescent taiga, eluding deluginal Burckle's abyssal excavation.

Offline baldusi

  • Senior Member
  • *****
  • Posts: 7437
  • Buenos Aires, Argentina
  • Liked: 1447
  • Likes Given: 4506
Re: Jim's HLV Spacecraft Design Thread
« Reply #224 on: 02/15/2012 07:38 PM »
We didn't imply that Ares V saved anything. We specifically are doubting the source of that data as highly partial and suspicious.
But let me explain something about the JWST. It works at less than 50 Kelvin. That's terribly cold to the level that they had to develop new bonding techniques for the carbon fibre structure because there was no known method that could work at that temperature.
They asked for a wavefront error that's never been achieved on such a big mirror. You can't polish a mirror at 50K. You have to calculate, do an approximation, freeze the mirror, measure your error, and repeat. To get equivalent performance, you'd have to at least have used a 5m mirror. I highly doubt there are freezing chambers that can go under 50 K and fit such a mirror. And even then, given the launching strains and all that, the trades where that it was actually easier to use smaller mirror with active optics to correct any deformations. Please remember, it would be on L2 and you can't correct any misalignment unless you have active mirror system.
So even if they had gone without the folding mechanism, they would most probably have used many small mirrors.
Again, I highly, but highly doubt that the folding mechanism was more than 3% to 5% of the total cost. That doesn't mean that you would have saved 3% to 5%. Because another solution would have had other problems. So you could have save something like 25% of 5% (i.e. 1.25%) if it was cheaper. But the truth is that we don't know.
Let's remember that the 8.5B includes operations and data analysis. So even if you had know that it was going to cost around 7.5B that would be close to 100M. Don't waste your time on this issue.

Offline JohnFornaro

  • Not an expert
  • Senior Member
  • *****
  • Posts: 9173
  • Delta-t is the salient metric.
  • Planet Eaarth
    • Design / Program Associates
  • Liked: 623
  • Likes Given: 333
Re: Jim's HLV Spacecraft Design Thread
« Reply #225 on: 02/18/2012 01:44 PM »
About trolling for 8086 chips, as an aside.

New tech is more expensive than old tech.  The payoff for chips, as one example, is that you get more transistors per dollar with the new chip, and this translates into more calculation power, therefore the amortized price of the calculation is less than what you got with the old chip.  Fine.

There's no question that an 80286 chip could never compete with a xeon chip doing an FEA computational analysis of some task, say combustion flow in the nozzle.  But one of the things that's happening these days, is that the common mindset is that one simply can't do any meaningful work improving launch costs, for example, without the latest and greatest tech.  But the lost truth is that plenty of work can be done without waiting for, or buying, the latest and greatest technology.

Somewhere on the forum is a thread which includes a short discussion on a multi-million dollar new computer and support contract.  Several posters observed that it was a pretty high annual cost per seat for tech support; maybe $10K/seat/year, IIRC.  It wasn't all that clear what NASA got for that money, however.  So what's the whine?

NASA is losing its way and its funding, in part because of high design costs.  The high costs are a partial function of a shallow insistance on new tech for the sake of new tech; losing sight of the work itself.

The typical bullying (see below) knee-jerk reaction is, "What? You don't want our Rocket Scientists to have the best tools available?" as if that is indeed the extent of the analysis necessary for getting work accomplished; and worse, as if the only choices I offer are between the latest and greatest and slide rules.  NASA's management would do well to learn something about "Appropriate Technology".

Again, the lost truth is that accomplishment is taking a back seat to the acquisition of new gadgetry.  The example of trolling for 8086's is an example of "for want of a penny a war was lost".

Part of a good spacecraft design would be in identifying the weak points in the design, such as chips, and ensuring that the supplies will be available over the life of the design.

As managerial styles go, bullying, which can be done quietly, is a longtime favorite; here's one example:

http://selenianboondocks.com/2010/06/agnep1/

Quote from: JonGoff
I have seen for years now at NASA how even the most clever engineers can be seduced or bullied into accepting terrible vehicle designs ...

He [Jon's boss] quietly but firmly cut me off and said:

"The answer" is Shuttle-C.”.

I understood from the tone of his voice that this decision wasn’t technical, it was strategic. Shuttle-C was based on propulsion hardware developed and controlled at MSFC. If the Mars program went forward, if this vehicle was developed, if Shuttle-C was baselined for its launch, then MSFC would be supporting it for many years to come.


That was from the past.  Today's example would be how the Orion capsule is simply not going to be made available to the ISS, if certain parties have their way.  The falsehood that Orion somehow competes with commercial crew is being offered as the rationale for this faulty course of action.  Other parties are arguing for the reduction of commercial crew funding, basically in order to "get even", I guess, with the opposing parties.  I suppose the biggest bully will win?

Crew safety at ISS is enhanced by redundancy; but this principle is abandoned so that the powerful parties can fight one another.  Cooperation is lost.  Costs will go up.  Accomplishment will be delayed.

Same old, same old.
Sometimes I just flat out don't get it.

Offline RyanC

  • Full Member
  • ****
  • Posts: 402
  • SA-506 Launch
  • Liked: 51
  • Likes Given: 9
Re: Jim's HLV Spacecraft Design Thread
« Reply #226 on: 02/18/2012 03:31 PM »
Since somehow my original post disappeared; here's a restatement and additional information:

In the 2000ies, NASA was trolling around for old bulk supplies of medical equipment to get Intel 8086 chips LINK

Then there's the case of the Solid Rocket Boosters.

The paper Reusable Solid Rocket Motor--Accomplishments, Lessons, and a Culture of Success by Moore and Phelps says that during a ten year period beginning in the mid 1990s, more than 100 materials used in the RSRM became obsolete -- aka got dropped by their manufacturers.

Yes, work arounds were found to keep the RSRM program, and by extension, the Shuttle program going; but it did consume a non-trivial amount of money for re-certification.

This never would have been a problem if the Shuttle had achieved it's program goal of rapid and routine spaceflight, because we'd have retired OV-099 through 104 by 1992-1994 and replaced them with the OV-2xx series; and the OV-2xx Orbiters would have in turn been retired by 2000-2002 and replaced with the OV-3xx series.

Offline Namechange User

  • Elite Veteran
  • Senior Member
  • *****
  • Posts: 7301
  • Liked: 0
  • Likes Given: 0
Re: Jim's HLV Spacecraft Design Thread
« Reply #227 on: 02/18/2012 03:37 PM »
Since somehow my original post disappeared; here's a restatement and additional information:

In the 2000ies, NASA was trolling around for old bulk supplies of medical equipment to get Intel 8086 chips LINK

Then there's the case of the Solid Rocket Boosters.

The paper Reusable Solid Rocket Motor--Accomplishments, Lessons, and a Culture of Success by Moore and Phelps says that during a ten year period beginning in the mid 1990s, more than 100 materials used in the RSRM became obsolete -- aka got dropped by their manufacturers.

Yes, work arounds were found to keep the RSRM program, and by extension, the Shuttle program going; but it did consume a non-trivial amount of money for re-certification.

This never would have been a problem if the Shuttle had achieved it's program goal of rapid and routine spaceflight, because we'd have retired OV-099 through 104 by 1992-1994 and replaced them with the OV-2xx series; and the OV-2xx Orbiters would have in turn been retired by 2000-2002 and replaced with the OV-3xx series.

Ryan,

You are being way overdramatic to try to just cast shuttle in a negative light. 

These type issues are not unique to Shuttle and were dealt with in any number of ways (and the reality is you have just scratched the surface) in many, many programs that exist in operations for years and years.

Are you similarly outraged at the B-52 program?
Enjoying viewing the forum a little better now by filtering certain users.

Offline RyanC

  • Full Member
  • ****
  • Posts: 402
  • SA-506 Launch
  • Liked: 51
  • Likes Given: 9
Re: Jim's HLV Spacecraft Design Thread
« Reply #228 on: 02/18/2012 04:18 PM »
Are you similarly outraged at the B-52 program?

In the early 1980s, the Gunner's CRT on the B-52 had a lifetime of about a decade in active use.

In the rare cases when they failed, they were replaced by new CRTs still in the original packaging and with production dates of 1961.

[talks with a former B-52 crewdog from the early 1980s]

Additionally in the 1990s, we retired all but the G and H models, in effect massively increasing the spare parts pile for the remaining B-52s.

The big problem was that the Shuttle was designed just as the microelectronics revolution was taking off in the 1970s -- the F-15 is prone to the same issues -- the USAF has been forced to trawl around to find the required chips for early F-15 flight control computers.

Offline RyanC

  • Full Member
  • ****
  • Posts: 402
  • SA-506 Launch
  • Liked: 51
  • Likes Given: 9
Re: Jim's HLV Spacecraft Design Thread
« Reply #229 on: 02/18/2012 04:24 PM »
These type issues are not unique to Shuttle and were dealt with in any number of ways (and the reality is you have just scratched the surface) in many, many programs that exist in operations for years and years.

The F-1A restart studies in the 1990s programmed in 42 specific material substitution recommendations.

Fourteen were due to regulations over the use of beryllium, cadmium and arsenic.

For example, the MK10A turbopump used an alloy which had a % of beryllium which exceeded the then-current OSHA standard.

Five were to replace obsolete materials like Inconel X750 and Hastelloy C.

Twenty-three were to replace materials that were stress corrosion susceptible, like 321 CRES, 17-7 PH Steel, 2024-T351 Aluminum and 7075 T6 Aluminum.

Offline Patchouli

  • Senior Member
  • *****
  • Posts: 4417
  • Liked: 189
  • Likes Given: 382
Re: Jim's HLV Spacecraft Design Thread
« Reply #230 on: 02/18/2012 04:27 PM »
About trolling for 8086 chips, as an aside.

New tech is more expensive than old tech.  The payoff for chips, as one example, is that you get more transistors per dollar with the new chip, and this translates into more calculation power, therefore the amortized price of the calculation is less than what you got with the old chip.  Fine.

There's no question that an 80286 chip could never compete with a xeon chip doing an FEA computational analysis of some task, say combustion flow in the nozzle.  But one of the things that's happening these days, is that the common mindset is that one simply can't do any meaningful work improving launch costs, for example, without the latest and greatest tech.  But the lost truth is that plenty of work can be done without waiting for, or buying, the latest and greatest technology.

Somewhere on the forum is a thread which includes a short discussion on a multi-million dollar new computer and support contract.  Several posters observed that it was a pretty high annual cost per seat for tech support; maybe $10K/seat/year, IIRC.  It wasn't all that clear what NASA got for that money, however.  So what's the whine?

NASA is losing its way and its funding, in part because of high design costs.  The high costs are a partial function of a shallow insistance on new tech for the sake of new tech; losing sight of the work itself.

The typical bullying (see below) knee-jerk reaction is, "What? You don't want our Rocket Scientists to have the best tools available?" as if that is indeed the extent of the analysis necessary for getting work accomplished; and worse, as if the only choices I offer are between the latest and greatest and slide rules.  NASA's management would do well to learn something about "Appropriate Technology".

Again, the lost truth is that accomplishment is taking a back seat to the acquisition of new gadgetry.  The example of trolling for 8086's is an example of "for want of a penny a war was lost".

Part of a good spacecraft design would be in identifying the weak points in the design, such as chips, and ensuring that the supplies will be available over the life of the design.

Choosing a chip that will be around for a long time is akin to looking into a crystal ball.
NASA did go right with the 8086 as it stuck around for a long time.
FYI a lot of embedded designs still even use 8bit controllers and there is some good arguments to using a simple processor for some things.
The 8051 chip for example is still manufactured as clones today.
Though they been largely replace by newer chips such as the At Mega328 there still is a good chance of a single chip 8051 clone being found inside a modern computer monitor or television.
It's sometimes used to drive the overlay display for the onscreen controls.
Though NASA did stick with the original 8086 vs move to newer clones or even single chip solutions.
Far more disturbing then using an old chip which is still common and can be found second sourced was the rumor NASA supposedly still used 8" floppy drives in some equipment.
I figure any 8" floppy would have long since been replace with 3.5" drives or even flash.


Are you similarly outraged at the B-52 program?

In the early 1980s, the Gunner's CRT on the B-52 had a lifetime of about a decade in active use.

In the rare cases when they failed, they were replaced by new CRTs still in the original packaging and with production dates of 1961.

[talks with a former B-52 crewdog from the early 1980s]

Additionally in the 1990s, we retired all but the G and H models, in effect massively increasing the spare parts pile for the remaining B-52s.

The big problem was that the Shuttle was designed just as the microelectronics revolution was taking off in the 1970s -- the F-15 is prone to the same issues -- the USAF has been forced to trawl around to find the required chips for early F-15 flight control computers.
I wonder if that was one reason they were so eager to retire the Apollo spacecraft as semiconductors went from SSI to LSI during the time frame of 1966 to 1975.
It's a really insane rate of change for just 9 years things seemed to have slowed down considerably since then.
Though the Shuttle did have core memory until 1985 and core was considered antiquated even in the mid 1970s.
« Last Edit: 02/18/2012 04:53 PM by Patchouli »

Offline Namechange User

  • Elite Veteran
  • Senior Member
  • *****
  • Posts: 7301
  • Liked: 0
  • Likes Given: 0
Re: Jim's HLV Spacecraft Design Thread
« Reply #231 on: 02/18/2012 04:30 PM »
Are you similarly outraged at the B-52 program?

In the early 1980s, the Gunner's CRT on the B-52 had a lifetime of about a decade in active use.

In the rare cases when they failed, they were replaced by new CRTs still in the original packaging and with production dates of 1961.

[talks with a former B-52 crewdog from the early 1980s]

Additionally in the 1990s, we retired all but the G and H models, in effect massively increasing the spare parts pile for the remaining B-52s.

The big problem was that the Shuttle was designed just as the microelectronics revolution was taking off in the 1970s -- the F-15 is prone to the same issues -- the USAF has been forced to trawl around to find the required chips for early F-15 flight control computers.

Excellent.  You are falling into my trap!

So now you admit there were workarounds, MTBF analysis, reliability studies, certified life lifits, etc.  These and other things, that you somewhat back-handedly, you have had to acknowledge we also had on Shuttle.  We got along just fine really.

You also do not acknowledge that there were upgrades to the systems, some due ot obsolescence and some for other reasons (that just reset the obsolescence clock). 

Technology moves on.  There is always something "new" or something "better".  Aircraft, spacecraft, ships or anything that is really a system-of-systems will always have to deal with this. 

It was not in any way unique to Shuttle and, again, we got along just fine and did the job expected of us in every way.  It will be present on future vehicles as well and I very much expect it to be handled in similar methods (which again were vast).  Just admit you hate Shuttle and are trying to apply a different filter to that vehicle and save us all some time.  ;)
« Last Edit: 02/18/2012 04:48 PM by OV-106 »
Enjoying viewing the forum a little better now by filtering certain users.

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 28616
  • Minnesota
  • Liked: 8585
  • Likes Given: 5600
Re: Jim's HLV Spacecraft Design Thread
« Reply #232 on: 02/18/2012 04:45 PM »
About trolling for 8086 chips, as an aside.

New tech is more expensive than old tech.  The payoff for chips, as one example, is that you get more transistors per dollar with the new chip, and this translates into more calculation power, therefore the amortized price of the calculation is less than what you got with the old chip.  Fine.

There's no question that an 80286 chip could never compete with a xeon chip doing an FEA computational analysis of some task, say combustion flow in the nozzle.  But one of the things that's happening these days, is that the common mindset is that one simply can't do any meaningful work improving launch costs, for example, without the latest and greatest tech.  But the lost truth is that plenty of work can be done without waiting for, or buying, the latest and greatest technology.

Somewhere on the forum is a thread which includes a short discussion on a multi-million dollar new computer and support contract.  Several posters observed that it was a pretty high annual cost per seat for tech support; maybe $10K/seat/year, IIRC.  It wasn't all that clear what NASA got for that money, however.  So what's the whine?

NASA is losing its way and its funding, in part because of high design costs.  The high costs are a partial function of a shallow insistance on new tech for the sake of new tech; losing sight of the work itself.

The typical bullying (see below) knee-jerk reaction is, "What? You don't want our Rocket Scientists to have the best tools available?" as if that is indeed the extent of the analysis necessary for getting work accomplished; and worse, as if the only choices I offer are between the latest and greatest and slide rules.  NASA's management would do well to learn something about "Appropriate Technology".

Again, the lost truth is that accomplishment is taking a back seat to the acquisition of new gadgetry.  The example of trolling for 8086's is an example of "for want of a penny a war was lost".

Part of a good spacecraft design would be in identifying the weak points in the design, such as chips, and ensuring that the supplies will be available over the life of the design.

Choosing a chip that will be around for a long time is akin to looking into a crystal ball.
NASA did go right with the 8086 as it stuck around for a long time.
FYI a lot of embedded designs still even use 8bit controllers and there is some good arguments to using a simple processor for some things.
The 8051 chip for example is still manufactured as clones today.
Though they been largely replace by newer chips such as the At Mega it of being found inside a modern computer monitor or television.
Though NASA did stick with the original 8086 vs move to newer clones or even single chip solutions.
...
Even if people stop building 8086s entirely, you can always just build one in an FPGA (there are fully rad-hard--not just rad-tolerant--FPGAs available these days). There are free IP 8086 "cores" available, even, though the cost of buying/licensing a non-free one would be absolutely trivial in this case. In fact, you could fit a whole bunch of those 8086s on a single FPGA. The nice thing about FPGAs, too, is that if you do want to upgrade to a more sophisticated core than the 8086 all you have to do is re-program it... you could do that even if it's billions of miles from Earth.

Also, with a rad-hard FPGA, you can program all (or almost all) the digital functions on the single unit, allowing you to effectively have your own single-chip solution.

EDIT:And for what it's worth, I disagree with the assertion that old tech is necessarily cheaper than modern tech. A modern microcontroller is a heck of a lot cheaper than a whole bunch of vacuum tubes or discrete transistors for just about anything digital and operates a lot faster and is more flexible as well (a bunch of digital inputs/outputs and often analog inputs and PWM/analog outputs thrown in for good measure) while operating with much less power and more reliable in most environments. Even when made in America. For about the price of a cup of coffee.
« Last Edit: 02/18/2012 05:05 PM by Robotbeat »
Chris  Whoever loves correction loves knowledge, but he who hates reproof is stupid.

To the maximum extent practicable, the Federal Government shall plan missions to accommodate the space transportation services capabilities of United States commercial providers. US law http://goo.gl/YZYNt0

Offline Patchouli

  • Senior Member
  • *****
  • Posts: 4417
  • Liked: 189
  • Likes Given: 382
Re: Jim's HLV Spacecraft Design Thread
« Reply #233 on: 02/18/2012 05:10 PM »
Even if people stop building 8086s entirely, you can always just build one in an FPGA (there are fully rad-hard--not just rad-tolerant--FPGAs available these days). There are free IP 8086 "cores" available, even, though the cost of buying/licensing a non-free one would be absolutely trivial in this case. In fact, you could fit a whole bunch of those 8086s on a single FPGA. The nice thing about FPGAs, too, is that if you do want to upgrade to a more sophisticated core than the 8086 all you have to do is re-program it... you could do that even if it's billions of miles from Earth.

Also, with a rad-hard FPGA, you can program all (or almost all) the digital functions on the single unit, allowing you to effectively have your own single-chip solution.

EDIT:And for what it's worth, I disagree with the assertion that old tech is necessarily cheaper than modern tech. A modern microcontroller is a heck of a lot cheaper than a whole bunch of vacuum tubes or discrete transistors and operates a lot faster and is more flexible as well (a bunch of digital inputs/outputs and often analog inputs and PWM/analog outputs thrown in for good measure) while operating with much less power and more reliable in most environments.

True FPGA's have probably made obsolescence less of an issue.
Even entire computers such as the C64 ,Apple II and even early Amigas have been replicated in FPGA.

The big issue with NASA is they were worried about software issues from unforeseen bugs.
They even avoided clones of the 8086 such as the NEC V30.

Though it should be noted hardware is only one part of the cost software and qualification also are considerable costs esp for something that is not mass produced in large numbers.

« Last Edit: 02/18/2012 05:17 PM by Patchouli »

Offline Robotbeat

  • Senior Member
  • *****
  • Posts: 28616
  • Minnesota
  • Liked: 8585
  • Likes Given: 5600
Re: Jim's HLV Spacecraft Design Thread
« Reply #234 on: 02/18/2012 05:37 PM »
Even if people stop building 8086s entirely, you can always just build one in an FPGA (there are fully rad-hard--not just rad-tolerant--FPGAs available these days). There are free IP 8086 "cores" available, even, though the cost of buying/licensing a non-free one would be absolutely trivial in this case. In fact, you could fit a whole bunch of those 8086s on a single FPGA. The nice thing about FPGAs, too, is that if you do want to upgrade to a more sophisticated core than the 8086 all you have to do is re-program it... you could do that even if it's billions of miles from Earth.

Also, with a rad-hard FPGA, you can program all (or almost all) the digital functions on the single unit, allowing you to effectively have your own single-chip solution.

EDIT:And for what it's worth, I disagree with the assertion that old tech is necessarily cheaper than modern tech. A modern microcontroller is a heck of a lot cheaper than a whole bunch of vacuum tubes or discrete transistors and operates a lot faster and is more flexible as well (a bunch of digital inputs/outputs and often analog inputs and PWM/analog outputs thrown in for good measure) while operating with much less power and more reliable in most environments.

True FPGA's have probably made obsolescence less of an issue.
Even entire computers such as the C64 ,Apple II and even early Amigas have been replicated in FPGA.

The big issue with NASA is they were worried about software issues from unforeseen bugs.
They even avoided clones of the 8086 such as the NEC V30.

Though it should be noted hardware is only one part of the cost software and qualification also are considerable costs esp for something that is not mass produced in large numbers.
Except that you can use already-qualified cores (which would be used in large numbers) on an FPGA. And the FPGAs themselves are used for commercial satellites and military applications so they are also relatively easy to come by... The cost isn't necessarily that high, either. I've seen small rad-hard FPGAs available for a little more than $1000 (that's University/amateur territory), though newer/better ones are probably in the $10,000+ range.

In general, though, you're right that the cost of the component itself is usually just a small consideration compared to the cost of the people who qualify and test it. And there's always Akin's Law #13:
"13. Design is based on requirements. There's no justification for designing something one bit "better" than the requirements dictate."
Chris  Whoever loves correction loves knowledge, but he who hates reproof is stupid.

To the maximum extent practicable, the Federal Government shall plan missions to accommodate the space transportation services capabilities of United States commercial providers. US law http://goo.gl/YZYNt0

Offline Chris Bergin

Re: Jim's HLV Spacecraft Design Thread
« Reply #235 on: 02/18/2012 06:02 PM »
This thread's run to the end of its lifetime.

Tags: